From the Center for Extreme Bionics, Massachusetts Institute of Technology; the Division of Plastic Surgery, Brigham and Women's Hospital; and the Department of Orthopaedics, University of Gothenburg.
Plast Reconstr Surg. 2019 Aug;144(2):218e-229e. doi: 10.1097/PRS.0000000000005864.
Traditional approaches to amputation are not capable of reproducing the dynamic muscle relationships that are essential for proprioceptive sensation and joint control. In this study, the authors present two caprine models of the agonist-antagonist myoneural interface (AMI), a surgical approach designed to improve bidirectional neural control of a bionic limb. The key advancement of the AMI is the surgical coaptation of natively innervated agonist-antagonist muscle pairs within the residual limb.
One AMI was surgically created in the hindlimb of each of two African Pygmy goats at the time of primary transtibial amputation. Each animal was also implanted with muscle electrodes and sonomicrometer crystals to enable measurement of muscle activation and muscle state, respectively. Coupled agonist-antagonist excursion in the agonist-antagonist myoneural interface muscles was measured longitudinally for each animal. Fibrosis in the residual limb was evaluated grossly in each animal as part of a planned terminal procedure.
Electromyographic and muscle state measurements showed coupled agonist-antagonist motion within the AMI in the presence of both neural activation and artificial muscle stimulation. Gross observation of the residual limb during a planned terminal procedure revealed a thin fibrotic encapsulation of the AMI constructs, which was not sufficient to preclude coupled muscle excursion.
These findings highlight the AMI's potential to provide coupled motion of distal agonist-antagonist muscle pairs preserved during below- or above-knee amputation at nearly human scale. Guided by these findings, it is the authors' expectation that further development of the AMI architecture will improve neural control of advanced limb prostheses through incorporation of physiologically relevant muscle-tendon proprioception.
传统的截肢方法无法复制对本体感觉和关节控制至关重要的动态肌肉关系。在这项研究中，作者提出了两种山羊的抗肌神经界面（AMI）模型，这是一种旨在改善仿生肢体双向神经控制的手术方法。AMI 的关键进步是在残肢内对天然神经支配的拮抗性肌肉对进行手术吻合。
在两只非洲俾格米山羊的后肢初次经胫骨截肢时，分别在每只动物的后肢上进行了一次 AMI 手术。每只动物还植入了肌肉电极和超声测微计晶体，分别用于测量肌肉激活和肌肉状态。对每个动物的 AMI 肌肉进行了纵向测量，以测量其拮抗肌的协同运动。作为计划的终末程序的一部分，对每个动物的残肢进行了大体评估，以评估纤维化情况。
肌电图和肌肉状态测量结果显示，在神经激活和人工肌肉刺激的情况下，AMI 内的拮抗肌协同运动。在计划的终末程序中对残肢进行大体观察发现，AMI 结构有一层薄薄的纤维性包膜，但不足以阻止肌肉协同运动。
这些发现强调了 AMI 在近人体比例的膝下或膝上截肢时提供保留下的远端拮抗肌对协同运动的潜力。基于这些发现，作者期望进一步开发 AMI 结构，通过整合具有生理相关性的肌肉-肌腱本体感觉，改善先进肢体假肢的神经控制。
